Bonjour,
Si j’achète un ic8 neuf aujourd’hui, est ce que j’aurai le problème de calibration ou le problème est désormais réglé ?
two weeks ago I bought Schwinn 800IC (new IE8 version). It looks like they fixed the power issue. My readings in Swift are accurate
Thank you!
I don’t understand the question. The power readings are transferred from the spin-bike to the zwift
I just wanted to know how you tested the accuracy.
1 Like
Ramp Test and perceived effort
1 Like
Hi everyone,
I just found this discussion very interesting due to the implication of product design which is something I love, beside triathlon…
So since a few weeks I bought Assioma Duo and have it mounted on my triathlon bike. The other day I decided to mount the pedal on my Schwinn IC8 and check considering the roumors around. On top I have been finding myself “flying” in Zwift and I got suspicious…
So here I am, as many of you with 50% off numbers and the IC8 heavily overestimating the power.
So I studied a bit the bike and technologies involved and after a day of testing and calculations here is my verdict:
-
I could find a relatively simple formula that can convert the IC8 power into Assioma power for any given combination of cadence and nob position (torque/resistance of the IC8). With this I can at least understand on my triathlon bike on the street where I am and relate to the indoor training values.
-
I could technically convert all my FIT files by readjusting according to this formula but I found that FitFileTools would not work since there you can only specify a single (multiplying) factor and you need also a constant to properly convert.
-
Physics works: both IC8 and Assioma data show that Power is a linear function of cadence and torque. But you need both of these value to calculate things correctly.
-
Since I am just training on Zwift and not doing official events or races, that is ok-ish… In the game it looks like I am a bullet but at least I have my references based on FTP tests. So when I define zones, they are relative to a wrong FTP but when I get to threshold (whatever number that is) my body feels right at threshold… So I might continue training on the IC8
-
It is a pity the company producing IC8 is not taking this up. It should be pretty trivial to correct from what I see. I am sure they are already working on a new bike with all this fixed and I would appreciate if they would not leave the community of the IC8 alone.
-
If you guys want to have the correct reading for Zwift there are 4 options: A) you buy yourself a proper trained like wahoo or garmin and you mount your bike there, B) you buy pedals with bluetooth powermeter like Assioma or Garmin and you connect Zwift to that source, C) we continue poking the manufacturer to fix the issue, D) if I have time I can think about a solution with a controller hack…
-
I checked the principles behind the power reading (yes, I am an analytics and engineering geek…) and the system is pretty simple with one twist. When you turn the nob, you gradually lower a component with magnets that gradually overlap with the rotating disk. The more they overlap, the more you get resistance. In my case I need 9.25 revolutions to go from one extreme (end of the run) to the other extreme (past 100 indicated resistance). Funny enough, to measure the position of the magnets with respect to the disk, the manufacturer does not use a distance or linear displacement detector but uses a magnetic system that measures the rotation at the connection point of the arm which hosts the magnets. So you push down by rotating the nob and the arm goes down rotating around a pin. The magnetic sensor measures this rotation (which is very small) and this is then used to associate it with the torque. So this is the critical parameter that it is used for the conversion in the firmware.
-
This special component that measure this rotation and sends the signal that is used to calculate the power is the AMS AS5600 chip which is great for such contactless applications (no friction = less degradation/wear with time). The datasheet of this component explains how to use it and calibrating it. So I understood that the procedure to calibrate the IC8 bike is nothing but the procedure to calibrate this chip. So you set the start angle/position, the end angle/position and then everything in between is divided in micro steps that can be read. The drawback of this procedure is that it is this sensor that has this limit of max. 3 calibrations (explicitly written in the datasheet). So it is nothing related to the Schwinn firmware but it is this chip (even more complex physics behind). If you exceed these 3 calibration options, the chip must be replaced.
-
Now that I dug into numbers conversions and physics, I understood that this recalibration procedure can only very marginally help. So unless your 100 reading appears more then 2 turns from the end of the run of the nob, then the calibration procedure will only have minimal impact. The difference between reality and IC8 power output is so much off that recalibration won’t correct it. So save the calibrations for when it is really needed!
What’s next from my side: I will probably dismount my Assioma and keep going with the standard pedals so that also my wife can continue training. I will try to see if I can figure out how to convert the power of my fit files from Zwift using the more complex formula I derived. I would like to think if there is a simple way to hack the system so that it sends the right power reading to Zwift.
I would be available if someone is interested in knowing more and share data so that I can verify if the formula I derived might apply across multiple devices.
Happy to help if I can.
Perhaps in the end of the day I will just buy a proper trainer. Pity for Schwinn, sometimes it takes so little to make something good, great. The product is good. I like to think that some engineer forgot a “2” somewhere in the calculations, this happens sometimes (either the 2 or the pi)…
Take care everyone and enjoy your rides!
Riccardo
For the records, I found that there is a program like GoldenCheetah that allows to implement the power correction with slope AND offset. So I could get reasonable match between the Assioma and the converted Schwinn IC8 power values.
Once you find the conversion rate for your bike calibration I guess you can apply this over and over again.
Via GoldenCheetah, you can re-export the file in FIT format and upload it on your platform of choice.
Of course this helps only for post-activity analysis while in Zwift you will always look awfully fast.
My FTP prior this mess was 325 W, after correction… 183 W!
Hi, this is the result of the analysis with some more graphical flavor.
As you can see the Schwinn reading is way off, whereas the Assioma original (dark green behind) is way lower… basically the real power peaks are as high as the lowest power indicated by the IC8… crazy!
Then I used three correction curves:
1 - Reconstructed 1: true_pwr = m1 x pwr_Ic8 + q1
2 - Reconstructed 2: true_pwr = m2 x pwr_Ic8 + q2 (so the same apprach but different coefficients)
3 - Scaled: true_pwr = m3*pwr_Ic8
As you can see scaling simply is not enough. I found the right coefficients now to convert the power considering the setting of my bike. So it could be done with all bikes I guess.
Ride on!
IC8 owner here. Just got the bike 2nd hand (like-new condition!) and having the same exact issues.
It’s been 2 years of complains so I’m guessing at this point it is a completely lost cause.
I got the new IC8 last week (Life Fitness IC8 Power Trainer)
I wasn’t expect the controller not working with Zwift
as smart bike I believe it should be more smart
wish I didn’t get it however when connect it to Zwift it show wahoo connected I am not sure this is normal or no
anyway I am still learn about IC8 and Zwift since I am new in both system
but look like I have to read too much as this is not easy like Wahoo or Garmin
I have a brand new IC4 bike and I recently got on Zwift did my first ride on Friday and I have not noticed the issues you all are going on about. I am on level 30 for resistance on the IC4 and was putting out 200-220wats. So I am not sure what to tell you folks. Setting up the IC4 to Zwift was a dream and so far going through resistance has been perfect.
How do you know that the resistance is perfect. I’m not saying that it is not but questioning the word perfect.
It’s flawless ? Is that better ? I have been testing it for awhile now and have a few friends with the same spin bike and all of us have had no issues and notice that resistance is… Perfect.
@Gerrie_Delport_ODZ point is: How do you know it is perfect?
In other words…how do you know if it is not reading 50 watts high/low? Perfect in this forum is you have a power meter reading exactly what you are putting out…i.e., a meter that matches your trainer output and no connection issues. Sounds like you have good connection…now show your power is correct.
Perfect for what? Perfect for enjoyable riding is wonderful if that’s the goal. Perfect for racing is something else since accuracy matters for that. Accuracy does not necessarily matter for having a fun time in the game, and that’s OK.
I have a good connection for sure. How would you like me to show the power is correct and I’ll try my best to do so.
Only way I know is to compare to a real power meter…for your IC8 that means power meter pedals. Or comparable to IRL with a power meter on a bike.
For fun, not a big deal…well really, no issue at all. If, however, you decide to race…it becomes an issue.
I’ll look into grabbing a set of power meter pedals. I may be able to find some within the biking community here to borrow and see. How does it effect the racing side of it ? I’m new to Zwift so I haven’t heard anything about having a Schwinn bike and racing
An accepted power meter (pedals/crank/etc) or smart trainer (wahoo/tacx/elite/saris/etc) is needed for a lot of racing. For some non-power trainers (Fluid2/KurtKinetic fluids/etc) where a known power curve is applied, one races under “ZPower”. Some races do not allow ZPower; you must have a known power meter/trainer or be disqualified.
The IC8, i believe, is not an accepted power output…Not sure you can even use for ZPower (plenty of others use it so they would need to chime in). The power curve/accuracy is just not known for the IC8. For Example: my wife can crank out 300 watts on a Peloton but can barely do 50 watts on my smart trainer (Saris H3). The Saris is an accepted device…the Peloton is wildly inaccurate.
For freeriding, group rides…you are ok. If you decide to race, an acceptable power meter is required. For high level racing, you will need an accepted, highly accurate smart trainer (short list) and a secondary power meter.
The cheapest way in, I believe, is single sided Assioma pedals. If you are in a bigger town/city…check for used devices.